Image pasting, or the creation of a composite image, is usually accomplished by having a system for acquiring images with a total field-of-view larger than the detector field-of-view (FOV). For applications such as full-spine imaging or long-legs imaging, the total coverage of anatomy (e.g., 60-120 cm) exceeds that of most current detectors and film-screen systems. Modifications of the current systems need to be designed to allow imaging a total coverage larger than the detector FOV. A problem with the current systems is that the digital images formed either need to be handled manually or do not give high quality images—making the pasted image less accurate.
Typical systems for image pasting use either: film-screen (FS) cassettes—either one long-film or multiple films in a large cassette; a series of stacked computed radiography (CR) plates—CR plates are also known as photostimulable phosphor (PSP) systems; or image intensifier (II) tubes.
Many FS systems have much lower image quality (IQ) than some detectors; are difficult to handle because several films must be manually placed into a cassette, and stored; have a lower dynamic range than some detectors; and cannot be post-processed, nor can they be quantitatively analyzed.
Typical CR systems tend to be difficult to handle because several PSP plates must be placed into a large, heavy, cumbersome, custom cassette; require extra time to insert and extract plates from the cassette; have a white band artifact at the region of overlap between the two plates due to attenuation; and have lower IQ than some detectors.
II systems tend to have poor IQ and introduce distortions such as pin-cushion effects. The detector can also cause representation errors during image pasting (this has been referred to as the Pisa effect, or s-distortion), have anisotropic resolution across the field (the resolution is highest in the center of the detector, but decreases toward the periphery), may be sensitive to vibrations during movement causing additional image noise called “microphonic” noise. The distortions take extra time and processing power to correct, and the system may need to take time after reaching a position before collecting an image in order to avoid some of the noise due to the vibrations. Longer processing and pauses could mean a longer exam time, which in turn may cause an increase in patient movement artifacts.
Also, images from some systems are typically not square, but rather round. With the round shape and distortion at the edges, only a small segment through the center of each image can be used for composing the pasted image (as the images move farther apart, the span of overlap decreases). Also, a larger number of images are required (15-60 images for round systems compared with 2-5 images with square systems—the largest round systems typically having a 30 to 40 cm field of view).
Image quality is important in image pasting with respect to both image noise and resolution. Systems with a lower detective quantum efficiency (DQE) cause images to exhibit high noise levels. Therefore, these systems require higher dose to image the patient in order to achieve comparable image quality. High dose is especially undesirable in pediatric cases, and especially, as in image pasting, when the subject of interest must be exposed multiple times in order to get a full image.
High resolution is desirable when viewing sharp details in orthopedic cases. For example, the spinal vertebrae end-plates and iliac crest are anatomies that require high resolution images to view. Sharp detail is also important when accurately joining two images to form a pasted image. A system for use in image pasting that can deliver a higher DQE is needed.
Typical image pasting systems also include image pasting algorithms which must search large portions of an image looking for an appropriate area of overlap. Such systems do not work as efficiently as possible because there is no frame of reference which suggests a possible starting point for the overlap. A system that can provide a starting point for an image pasting algorithm to locate the overlap is desirable.